Scholarlink

<p>To address the low efficiency and local-optimum tendency caused by the dense distribution of weld points and the complex topology of precast beam reinforcement frameworks, this study proposes an improved ant colony optimization (IACO) algorithm combining elite strategy and adaptive mechanisms. A 3D coordinate model of the weld points is established, and the welding path planning problem is transformed into a three-dimensional traveling salesman problem (TSP). Adaptive adjustments are introduced in path construction, pheromone updating, and elite retention to better balance global search and local exploitation. Simulation experiments were carried out in single-group and five-group stirrup scenarios. In the single-group case, the algorithm converged quickly and the planned path was verified through robotic welding experiments. In the five-group scenario, IACO was compared with standard ACO, GA, and PSO. It converged to the optimal path length of 18,602.93&#xa0;mm at the 4th iteration, taking only 2.0567s, which highlights its outstanding computational efficiency and performance.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Intelligent path planning for robotic welding of precast beam reinforcement frameworks using an improved ant colony optimization

  • 期刊论文

  • 发表日期：

    2026年6月27日

查看全文
  • Songwei Li,
  • Kexin Feng,
  • Shiyu Guan,
  • Wei Liang,
  • Xian Cui,
  • Ting Lei

摘要

To address the low efficiency and local-optimum tendency caused by the dense distribution of weld points and the complex topology of precast beam reinforcement frameworks, this study proposes an improved ant colony optimization (IACO) algorithm combining elite strategy and adaptive mechanisms. A 3D coordinate model of the weld points is established, and the welding path planning problem is transformed into a three-dimensional traveling salesman problem (TSP). Adaptive adjustments are introduced in path construction, pheromone updating, and elite retention to better balance global search and local exploitation. Simulation experiments were carried out in single-group and five-group stirrup scenarios. In the single-group case, the algorithm converged quickly and the planned path was verified through robotic welding experiments. In the five-group scenario, IACO was compared with standard ACO, GA, and PSO. It converged to the optimal path length of 18,602.93 mm at the 4th iteration, taking only 2.0567s, which highlights its outstanding computational efficiency and performance.